VA testing is essential to evaluate the visual function. Its definition refers to the ability to recognize stimuli with high contrast and subtending a known angle [8]. According to these principles, different methods have been designed, taking into account the age and capacity of the patients.
For literate people, letters are a good method of recognition; HOTV, Snellen and ETDRS are the most frequently used charts in clinical practice. The Snellen chart has a series of drawbacks such as a different number of letters at each level and different vertical and horizontal distance between stimuli [9]. These limitations have been largely overcome with the LogMAR acuity charts, as the ETDRS, which is considered the gold standard tool for the measurement of VA both in clinical practice and in research[10-11].
HOTV optotype, composed of the letters H, O, T and V, is of great value for testing children at the age of starting reading, around 4-5y [12-13]. All these letters are symmetric, which avoid the common mistake of inverting letters.
In children and patients who do not know the letters, symbols may be useful. LEA, tumbling E and Landolt C are the most common ones, but these last two, that represent visual resolution acuity, rely on the child's spatial perception, and sometimes this ability is not well developed until the age of 4 y [14].
LEA is widely used in clinics and research for children under 4 y. It was designed to avoid some cultural barriers using common pictures (square, circle, house, heart), and speech problems or shyness with the option of matching figures[5,15]. Many studies have compared LEA symbols and letter optotypes in children, most of them showing good agreement [16-18]. However, the question about comparison between symbols and letters is still open, because some works addressed a tendency to higher VA values when measuring with LEA [19-20], while others turn to perform on the contrary [21]. Comparing letters and symbols in adults could clarify the effect of the stimulus design in VA, avoiding the bias of attention in children.
In our sample, values obtained by the different tests substrate an excellent agreement. The difference between the highest and lowest means is of 0.09 LogMar, which represents less than one line of VA level in common optotypes, which is clinically acceptable. Nevertheless, some points could be noted related to the differences.
Our patients obtained slightly lower results in symbol charts than in letters. In the case of printed tests, LEA was harder to recognize than the letters of ETDRS, and in DIVE tests, DIVE symbols obtained worse VA values than HOTV. The difference could be due to familiarity with stimuli, being letters in adulthood more common than draws. However, all these four tests were based in recognition acuity and their designs followed exactly the same principles. Therefore, results should be comparable in terms of recognition.
Photometric conditions vary across VA tests and can lead to a lack of standardization. In this sense, digital devices have more control of luminance conditions and offer a wider spectrum of contrasts. Livingstone et al. reported better accomplish of international photometric standards with iPad tablet devices than with retro illuminated ETDRS charts, in a standard clinical use [22]. Higher contrast levels in digital devices has been pointed as the reason for higher visual acuities [23]. This could explain our slightly higher VA values obtained by DIVE tests. However, it should be noted that luminance levels cannot be easily assessed in printed test, therefore direct comparisons were not done.
Pixelisation of digital screens appeared as a limitation to offer small visual stimuli and therefore to test high visual acuities. However, the availability of high resolution screens allow to display enough stimuli sizes to assess VA, even in short distances to the viewer [24]. Actually, clinical studies have demonstrated no relevant effect of pixelisation when comparing digital and printed charts [25].
Another source of bias in digital screens can be glare, that results in significant lower VA values in studies with tablets without anti-glare [25]. In our experience this factor was not present, since VA values in DIVE were slightly higher than obtained in printed tests. It could be due to the relative control of the external light sources in a clinical setting.
DIVE used one isolated figure on the screen surrounded by four bars. This method has been called contour interaction. It differs from the crowding effect that is the degradation of VA when a target is flanked by similar stimuli [26]. Crowding effect was present in printed charts as ETDRS and LEA, while in DIVE testing there was only contour interaction, which seems to have less influence in VA.
Since the arrival of portable digital technology, there have been an increasing number of studies that compare traditional versus digital visual tests. Most of them highlight the good correlation between gold standards and these emerging applications, displayed on tablets and smartphones [27-31].
In our study, DIVE tests demonstrated a high correlation with printed tests widely used, which make it a suitable option to measure VA. In addition to the validity, DIVE offers remarkable advantages. The control of stimulus design in terms of size, luminance or shape, can adapt VA testing to the patient’s characteristics. Results are recorded on the store data, allowing better follow up and automatic comparison with normal values. The random order of presentation avoids chart memorization and the logarithmic intervals adapted to the patient’s responses improve accuracy.
The small number of patients comprised may limit the generalizability, specifically the absence of children.
For better comparison purposes it would be interesting to compare repeatability, testing time and user’s satisfaction level with the different VA tests.